Download hip fractures in men with prostate cancer treated with

0022-5347/04/1726-2208/0
THE JOURNAL OF UROLOGY®
Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 172, 2208 –2212, December 2004
Printed in U.S.A.
DOI: 10.1097/01.ju.0000143930.73016.c6
HIP FRACTURES IN MEN WITH PROSTATE CANCER TREATED
WITH ORCHIECTOMY
PAUL W. DICKMAN,* JAN ADOLFSSON, KENT ÅSTRÖM
AND
GUNNAR STEINECK
From the Departments of Oncology and Pathology (Clinical Cancer Epidemiology) (PWD, GS) and Medical Epidemiology and Biostatistics
(PWD), and Oncologic Center (JA), Karolinska Institutet, Stockholm, and Department of Orthopedics (KA), Täby Närsjukhus, Sweden
ABSTRACT
Purpose: Androgen deprivation therapy increases the risk of osteoporosis related fractures.
This issue is of increasing importance in men with prostate cancer as increasingly more undergo
androgen deprivation therapy and therapy is administered sooner following diagnosis. Data
directly addressing the long-term fracture risk in men diagnosed with prostate cancer are
limited.
Materials and Methods: Using population based registries in Sweden we studied the incidence
of hip fractures in 17,731 men diagnosed with prostate cancer from 1964 to 1996 who were
treated with bilateral orchiectomy within 6 months of diagnosis. The fracture incidence was
compared to the incidence in 43,230 men diagnosed with prostate cancer but not treated with
orchiectomy and in 362,354 of similar age who were randomly selected from the general population.
Results: Men treated with orchiectomy were at increased risk for hip fracture. The estimated
relative risk comparing men who underwent orchiectomy to population controls was 2.11 (95% CI
1.94 to 2.29) for femoral neck fractures and 2.16 (95% CI 1.97 to 2.36) for intertrochanter
fractures. An increased risk of hip fracture was observed as early as 6 months after orchiectomy
and the relative risk remained fairly constant up to 15 years following orchiectomy.
Conclusions: Hip fracture risk increases almost immediately following orchiectomy and the
excess risk persists for at least 15 years. This side effect should be considered when assessing the
merits of androgen deprivation therapy, particularly in symptom-free men diagnosed with
localized prostate cancer. Measures to prevent osteoporosis should be considered in men undergoing androgen deprivation therapy.
KEY WORDS: prostate, androgen antagonists, fractures, orchiectomy, prostatic neoplasms
Osteoporosis related fractures are an important complication of androgen deprivation therapy in men with prostate
cancer.1, 2 Androgen deprivation can be achieved surgically
by bilateral orchiectomy or medically, eg by treatment with
gonadotropin-releasing hormone agonists. The relationship
between hypoganadism and osteoporosis has been appreciated for decades and a decreased incidence of femur neck
fracture has been observed in women on hormone replacement therapy but only recently has research focused on osteoporosis and its consequences in men undergoing androgen
deprivation therapy.3
Androgen deprivation therapy is associated with decreased
bone mineral density in men diagnosed with prostate cancer1
as well as in young men castrated following sexual offenses.4
Men treated with androgen deprivation therapy for prostate
cancer have been shown to be at increased risk for bone
fracture, although previous studies have been relatively
small.5–11
The incidence of localized prostate cancer is increasing in
many countries, partly due to the effect of prostate specific
antigen screening. Androgen deprivation therapy, previously
used primarily in men with metastatic prostate cancer, is
being administered in an increasing proportion of men diagnosed with nonmetastatic prostate cancer, who have relatively long life expectancy, and therapy is commencing
sooner following diagnosis. Therefore, total time spent with
castrate levels of testosterone in men diagnosed with prostate cancer is increasing. Thus, the prevention of osteoporosis
related fractures, which are associated with considerable
morbidity and mortality, is an increasingly important issue.
Using population based registries available in Sweden we
examined the incidence of various bone fractures in 17,731
men diagnosed with prostate cancer treated with bilateral
orchiectomy. The fracture incidence was compared to the
incidence in 43,230 men diagnosed with prostate cancer but
not treated with orchiectomy and in 362,354 of similar age
who were randomly selected from the general population. We
estimated the increase in fracture risk associated with androgen deprivation therapy and studied how fracture risk
depends on time since the commencement of therapy, calendar period and patient age.
METHODS
The unique national registration number assigned to all
Swedish residents is used to index almost all registries in Sweden, including the cause of death registry, cancer registry, registry of total population and hospital inpatient registry.12 The
population based Swedish Cancer Registry commenced in 1958.
Notification of all newly diagnosed cancers is mandatory for
pathologists/cytologists and physicians. Approximately 96% of
all incident cancers are reported to the Swedish Cancer Registry and 98% of those reported are histologically verified.13 The
Swedish Cancer Registry does not record information on stage
or treatment. Individual information on inpatient care has been
documented county-wise in the inpatient register since 1964,
Accepted for publication July 2, 2004.
Supported by the King Gustav V jubilee fund.
* Correspondence: Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Box 281, 171 77 Stockholm, Sweden
(telephone: ⫹46 8 5248 6186; FAX: ⫹46 8 314 975; e-mail:
[email protected]).
2208
2209
HIP FRACTURES AND PROSTATE CANCER
when 6 of the 26 Swedish counties provided data. Of the 26
counties 20 reported all inpatient care to the register in 1984
and coverage has been nationwide since 1987.14
Using the cancer register we identified all men diagnosed
with prostate cancer between 1964 and 1996. For each man
diagnosed with prostate cancer we randomly selected from
the general population using the Registry of Total Population
5 of the same age residing in the same county and without a
diagnosis of prostate cancer. The study was restricted to
men diagnosed at a time when their county provided data
to the inpatient registry since information on exposure (bilateral orchiectomy) and outcomes (fractures of the spine,
pelvis, femoral neck, intertrochanter, thigh, lower leg, eg
tibia or fibia, ankle, skull, skull base and face) was obtained
by matching with the hospital inpatient registry.
Individuals were considered to enter the study (be at risk)
6 months following diagnosis of prostate cancer. Men diagnosed with prostate cancer who underwent bilateral orchiectomy within 6 months of diagnosis were classified into the
orchiectomy group and men diagnosed with prostate cancer
who were not treated with bilateral orchiectomy were classified into the no orchiectomy group. A third group contained
population controls. Men who underwent bilateral orchiectomy 6 months or more following prostate cancer diagnosis
were excluded from study. All men were followed until death,
emigration, first diagnosis of the specific bone fracture or
December 31, 1996. Each fracture was studied separately
and only the first fracture of each bone was considered. We
focused primarily on fractures of the femoral neck and intertrochanter since such fractures are relatively common and
usually result in hospital admission.
We estimated the incidence proportion (cumulative incidence) for each bone fracture using the Kaplan-Meier
method. We then modeled fracture incidence rates using
Poisson regression to estimate incidence rate ratios (IRRs)
and associated 95% CIs in each of the 2 groups of men
diagnosed with prostate cancer compared to population controls. Estimates were adjusted for calendar period of diagnosis (grouped as 1960 to 1983, 1984 to 1991 and 1992 to 1996),
time since entry (0 to 1, 1 to 2, 2 to 3, 3 to 4, 4 to 5, 5 to 7, 7
to 10, 10 to 14 and 15 or greater years) and attained age (0 to
64, 65 to 69, 70 to 74, 75 to 79, 80 to 84 and 85 or older). We
used interaction terms in the model to investigate how the
effect of exposure varied according to time since entry, calendar period of diagnosis and attained age.
We also estimated the relative excess risk15 as a means of
comparing fracture risk between the orchiectomy and no
orchiectomy groups. Excess risk was estimated as the difference between the estimated fracture incidence rate in the
specific group minus the estimated fracture incidence rate in
population controls. We then calculated relative excess risk
as the ratio of the 2 estimated excess risks.
Evaluation of model goodness of fit was performed using
the deviance statistic, the study of deviance residuals and the
study of interaction terms. All analyses were performed using the SAS system, version 8.02 (SAS Institute, Carey,
North Carolina). We have a large material, meaning that
even relatively small differences become statistically significant, so we do not emphasize the statistical significance of
our results or report p values. Rather, we present 95% CIs as
measures of statistical precision. Data from the various registries were matched at the Swedish National Board of
Health and Welfare, and were provided to us without personal identifiers. As such, ethics committee approval was not
required.
RESULTS
The study included 17,731 men with prostate cancer
treated with bilateral orchiectomy within 6 months of diagnosis (table 1). Median time from diagnosis to orchiectomy in
these men was 40 days (mean 51). The study also included
43,230 men diagnosed with prostate cancer but not treated
with bilateral orchiectomy and 362,354 population controls.
The number of orchiectomies performed yearly varied with
calendar time, reflecting trends in the preferred type of androgen deprivation therapy (fig. 1). Figure 1 underestimates
the total number of orchiectomies performed in Sweden since
men were required to survive 6 months or longer following
diagnosis to enter our study. When assessing trends in the
number of orchiectomies with time, it must be considered
that not all counties provided data to the inpatient registry
prior to 1987.
Table 2 shows estimated incidence rate ratios for each of
the 2 groups of men diagnosed with prostate cancer compared to controls. We observed an approximately double risk
of fractures of the spine, pelvis, femoral neck, intertrochanter, thigh and lower leg in men treated with orchiectomy
compared to population controls. Men diagnosed with prostate cancer who did not undergo orchiectomy were also at
increased risk for these bone fractures compared to population controls, although the increase in risk was not of the
same magnitude as in men treated with orchiectomy.
Fractures of the femoral neck and intertrochanter were the
most common fractures observed during followup. Table 3
shows incidence rate ratios for these 2 fractures estimated
separately for various time intervals since entry. An increased risk was observed from the first interval, which covers the period of 6 to 18 months following diagnosis, and
relative risks remained at a similar level throughout followup. This feature could be seen in plots of cumulative
incidence (figs. 2 and 3), although cumulative incidence estimates were not adjusted for potential confounding factors.
The incidence proportion of femoral neck fractures 10 years
following diagnosis was 12% in men treated with orchiectomy
and 5% in men without a diagnosis of prostate cancer (fig. 2).
The corresponding values for intertrochanter fractures were
9% and 4% (fig. 3).
The estimated relative risks of femoral neck and intertrochanter fractures decreased with increasing patient age (table 4). This is because the baseline fracture incidence rate
increased with age (table 5). Because there were fewer fractures in the younger age groups, the estimated relative risks
reported in other tables is weighted toward the estimates for
the older age groups. The ratios of excess risks (ie relative
excess risks) were similar for each age group (table 5). There
was some evidence that the relative risks in men treated with
and without orchiectomy increased with calendar time (table
6). Study of the underlying incidence rates showed that incidence increased with time in men treated with orchiectomy,
remained stable in men treated without orchiectomy and
decreased in population controls (data not shown).
DISCUSSION
We found that men treated with orchiectomy as a means of
androgen deprivation were at increased risk for hip fracture.
Increased risk was observed as early as 6 months after orchiectomy and the relative risk remained fairly constant up
to 15 years following orchiectomy. The relative risk was
higher in younger than in older men. Restricting analysis to
men diagnosed in 1992 or later did not alter the findings.
Fractures can be broadly classified as osteoporotic or
pathological, or due to major trauma. We do not have information on the underlying cause of the observed fractures and
in particular we could not identify pathological fractures,
which are caused by tumor growth in the bone. There is no
reason to believe that men diagnosed with prostate cancer
are at higher risk for major trauma than population controls,
and so including fractures due to major trauma would have
attenuated relative risk estimates. During the late 1970s and
early 1980s, before the era of prostate specific antigen screen-
2210
HIP FRACTURES AND PROSTATE CANCER
TABLE 1. Characteristics of individuals in study
Characteristics
No. individuals
No. at risk for femoral fracture after
(yrs):
1
2
5
10
15
No. age at entry (%):
64 or Younger
65–69
70–74
75–79
80–84
85 or Older
Age at entry:
Mean
Median
SE
Range
No. entry yr (%):
1964–1983
1984–1991
1992–1996
No. fractures during followup:
Spine
Pelvis
Femur neck
Intertrochanter
Thigh
Lower leg (tibia, fibia)
Ankle
Skull
Skull base
Face
Totals
Person-yrs at risk:
Mean
Median
SE
Range
Yrs from entry to death during followup:
Mean
Median
SE
Range
No.
Yrs from entry to first bone fracture:
Mean
Median
SE
Range
Orchiectomy
No Orchiectomy
Controls
17,731
43,230
362,354
13,443
9,909
4,029
577
30
33,508
25,682
11,292
2,536
498
313,113
267,309
156,752
50,366
12,041
1,583 (9)
2,463 (14)
4,193 (24)
4,720 (27)
3,350 (19)
1,422 (8)
6,797 (16)
7,380 (17)
10,381 (24)
9,478 (22)
6,245 (14)
2,949 (7)
46,984 (13)
56,919 (16)
85,671 (24)
84,155 (23)
59,216 (16)
29,409 (8)
75
75
0.06
44–97
73
73
0.04
38–98
74
74
0.01
38–101
2,547 (14)
11,139 (63)
4,045 (23)
9,373 (22)
16,015 (37)
17,842 (41)
82,270 (23)
162,945 (45)
117,139 (32)
195
122
604
491
37
41
21
7
14
29
253
181
920
701
45
61
39
12
23
66
2,833
2,318
9,956
7,772
407
693
475
155
258
746
1,561
2,301
25,613
3.38
2.51
0.02
0–25
3.59
2.58
0.01
0–29
2.80
2.05
0.02
0–18
13,469
3.46
2.42
0.02
0–29
22,044
5.31
4.23
0.01
0–30
3.26
2.60
0.07
0–17
5.47
4.41
0.13
0–23
4.95
3.93
0.03
0–26
TABLE 2. Estimated IRRs of fracture risk in men diagnosed with
prostate cancer (treated with and without orchiectomy) vs men
without prostate cancer
No Orchiectomy
IRR
95% CI
Orchiectomy
IRR
95% CI
Spine
1.26
1.10–1.43
2.34
2.02–2.71
Pelvis
1.15
0.99–1.34
1.79
1.49–2.15
Femur neck
1.35
1.26–1.44
2.11
1.94–2.29
Intertrochanter
1.33
1.23–1.44
2.16
1.97–2.36
Thigh
1.54
1.13–2.10
3.52
2.49–4.95
Tibia, fibia
1.18
0.90–1.53
2.16
1.57–2.97
Ankle
0.99
0.71–1.37
1.78
1.14–2.77
Skull
0.97
0.54–1.76
1.52
0.71–3.26
Skull base
1.24
0.81–1.91
1.91
1.11–3.29
Face
1.18
0.91–1.51
1.34
0.93–1.95
Skull ⫹ face
1.21
0.98–1.49
1.54
1.15–2.06
Estimates are adjusted for attained age, time since entry, and calendar
period at diagnosis.
FIG. 1. Number of orchiectomies in men in study for each calendar
year from 1965 to 1996.
ing, approximately a quarter of men diagnosed with prostate
cancer in Sweden had metastatic disease at diagnosis.16 Androgen deprivation is a common therapy in men diagnosed
with metastatic prostate cancer and we would expect that the
group treated with orchiectomy would have contained a
larger proportion with metastatic disease at diagnosis compared to the group that did not undergo orchiectomy. As
such, we might have expected a higher incidence of patho-
2211
HIP FRACTURES AND PROSTATE CANCER
TABLE 3. IRRs estimated separately for each period since entry at 6
months following diagnosis
Yrs Since Entry
No Orchiectomy
IRR
95% CI
Femur neck:
0–1
1.45
1.25–1.68
1–2
1.39
1.18–1.64
2–3
1.42
1.19–1.71
3–4
1.08
0.86–1.36
4–5
1.32
1.04–1.67
5–7
1.41
1.17–1.71
7–10
1.32
1.07–1.63
10–14
1.20
0.90–1.59
15 or Greater
1.32
0.77–2.26
Intertrochanter:
0–1
1.74
1.48–2.06
1–2
1.42
1.18–1.71
2–3
1.07
0.84–1.35
3–4
1.22
0.95–1.57
4–5
1.22
0.93–1.59
5–7
1.21
0.97–1.51
7–10
1.13
0.88–1.45
10–14
1.32
0.97–1.81
15 or Greater
1.64
0.98–2.77
Estimates were obtained using interaction terms
model and are adjusted for attained age and calendar
Orchiectomy
IRR
95% CI
1.81
2.07
2.41
2.18
2.55
2.10
2.17
1.27
5.61
1.50–2.19
1.71–2.52
1.96–2.96
1.72–2.77
1.97–3.28
1.66–2.66
1.65–2.87
0.68–2.37
1.40–22.6
2.38
1.95–2.90
1.84
1.46–2.32
2.27
1.80–2.87
2.73
2.13–3.50
1.70
1.22–2.37
2.01
1.55–2.61
1.90
1.39–2.60
2.96
1.90–4.62
2.86
0.40–20.4
in Poisson regression
period at diagnosis.
FIG. 3. Incidence proportion of intertrochanter fractures in each of
3 exposure groups as function of time since start of followup (6
months after diagnosis).
TABLE 4. IRRs and 95% CIs estimated separately according to
attained age
Attained Age
No Orchiectomy
IRR
95% CI
Orchiectomy
IRR
Femur neck:
0–64
3.71
2.36–5.81
7.63
65–69
2.05
1.44–2.92
6.66
70–74
1.35
1.09–1.68
2.93
75–79
1.44
1.24–1.67
2.52
80–84
1.27
1.12–1.45
1.91
85 or Older
1.26
1.12–1.41
1.63
Intertrochanter:
0–64
2.81
1.68–4.72
10.80
65–69
1.75
1.22–2.53
4.39
70–74
1.57
1.23–2.00
3.65
75–79
1.26
1.06–1.50
2.28
80–84
1.29
1.11–1.50
1.77
85 or Older
1.24
1.08–1.41
1.89
Estimates were obtained using interaction terms in Poisson
model and are adjusted for calendar period and time since entry.
FIG. 2. Incidence proportion of femoral neck fractures in each of 3
exposure groups as function of time since start of followup (6 months
after diagnosis).
logical fractures in the group of men who underwent orchiectomy. However, men diagnosed with metastatic prostate
cancer have a poor prognosis, and so most pathological fractures would occur within a year or 2 following diagnosis. By
excluding the first 6 months of followup following diagnosis
we excluded men in whom tumor was diagnosed as a result of
a fracture as well as men who experienced fracture within 6
months of diagnosis. We observed approximately constant
relative risks throughout followup, suggesting that our estimates were not heavily influenced by the occurrence of
pathological fractures. Skeletal metastases are more common in the vertebrae and pelvis than in the hip, so that if our
estimates were heavily influenced by pathological fractures,
we would have expected larger relative risks for these bones,
which we did not observe.
The methods of diagnosing and treating prostate cancer
have changed with time. We observed similar patterns of
relative fracture risk when analyses were restricted to individuals diagnosed in 1992 or later, eg relative risk estimates
were similar for each time interval since entry. However, the
absolute fracture risk during this period was slightly higher
in men who underwent orchiectomy and slightly lower in
population controls, leading to slightly higher estimates of
relative risk. During the 1990s compared to earlier years a
smaller proportion of men diagnosed with prostate cancer
95% CI
3.92–14.9
4.55–9.75
2.27–3.79
2.12–2.99
1.64–2.22
1.40–1.89
5.94–19.8
2.82–6.82
2.78–4.80
1.87–2.78
1.48–2.12
1.62–2.21
regression
had metastatic disease at diagnosis. As such, we would have
expected the influence of pathological fractures on the relative fracture risk comparing men with and without orchiectomy to be smaller during this period. That is, we would have
expected the estimated relative risks to be larger for this
period, which is exactly what we observed. Estrogen therapy,
which is thought to protect against osteoporosis,17 was commonly used as a means of androgen deprivation during 1964
to 1983, so that we might have expected patients diagnosed
with prostate cancer to be at decreased fracture risk compared to population controls during this period, which we did
not observe.
The observed difference in fracture risk between, for example, men treated with orchiectomy and population controls cannot be completely attributed to osteoporotic fractures secondary to orchiectomy. Other factors that affect
fracture risk, the probability of being treated as an inpatient
when a fracture is diagnosed, the probability that a diagnosed fracture is recorded in the inpatient registry or selection into the groups that we studied should be considered. It
is reasonable to believe that most individuals with a hip
fracture have the fracture diagnosed and are admitted to the
hospital for surgery. On the other hand, vertebral fractures
are not always diagnosed and even when a vertebral fracture is diagnosed, the patient is not always admitted to the
hospital. For most fracture types including those of the hip,
we believe that the sensitivity and specificity of recording
fractures was similar in the 3 study groups. Therefore, we
argue that our relative risk estimates should not be substantially biased.
2212
HIP FRACTURES AND PROSTATE CANCER
TABLE 5. Fractures, person-years at risk and estimated incidence rates for femoral neck and intertrochanter fractures stratified by
attained age
Controls
Attained Age
No.
Fractures
Person-Yrs
at Risk
No Orchiectomy
Rate/10,000
Person-Yrs
No.
Fractures
Person-Yrs
at Risk
Rate/10,000
Person-Yrs
Orchiectomy
No.
Fractures
Femoral neck:
0–64
64
132,709
5
27
15,118
18
65–69
179
216,856
8
37
21,854
17
70–74
741
377,201
20
92
34,708
27
75–79
1,713
472,572
36
201
38,563
52
80–84
2,792
409,415
68
252
29,100
87
85 or Older
4,005
297,026
135
311
18,408
169
Intertrochanter:
0–64
59
132,685
4
19
15,140
13
65–69
192
216,849
9
34
21,870
16
70–74
528
377,435
14
76
34,740
22
75–79
1,411
473,215
30
144
38,662
37
80–84
2,141
411,140
52
194
29,236
66
85 or Older
3,101
298,832
104
234
18,535
126
* Estimated excess risk ratio in men with prostate cancer who did vs did not undergo orchiectomy.
TABLE 6. IRRs and 95% CIs estimated separately for each
calendar period
Calendar Period
No Orchiectomy
IRR
95% CI
Orchiectomy
IRR
Femur neck:
1964–1983
1.22
1.08–1.37
2.00
1984–1991
1.37
1.25–1.52
2.05
1992–1996
1.56
1.34–1.82
2.52
Intertrochanter:
1964–1983
1.25
1.08–1.44
1.84
1984–1991
1.30
1.17–1.45
2.17
1992–1996
1.52
1.27–1.81
2.52
Estimates were obtained using interaction terms in Poisson
model and are adjusted for attained age and time since entry.
95% CI
1.64–2.44
1.85–2.27
2.04–3.11
1.44–2.36
1.94–2.42
1.99–3.19
regression
Androgen deprivation therapy is being administered in an
increasing proportion of men diagnosed with localized prostate cancer who have relatively long life expectancy and
therapy is commencing sooner following diagnosis. Our finding that hip fracture risk increases almost immediately following orchiectomy and excess risk persists for at least 15
years suggests that the increased risk of hip fracture should
be considered when assessing the merits of androgen deprivation therapy, particularly in symptom-free men diagnosed
with localized prostate cancer. The potential benefits of improved survival associated with androgen deprivation therapy in men with localized prostate cancer must be weighed
against the potential side effects, including an increased risk
of osteoporosis and osteoporotic fractures. Measures to prevent osteoporosis should be considered in men undergoing
androgen deprivation therapy and proposed prevention strategies include lifestyle modifications, supplemental calcium
and vitamin D,2 and decreased vitamin A.18 Recent evidence
suggests that biphosphonates may be beneficial for preventing bone loss during androgen deprivation therapy, although
further studies in this area are required.3, 19
Elin Eriksson and Emma Jaensson assisted with data
management and analysis.
REFERENCES
1. Ross, R. W. and Small, E. J.: Osteoporosis in men treated with
androgen deprivation therapy for prostate cancer. J Urol, 167:
1952, 2002
2. Smith, M. R.: Diagnosis and management of treatment-related
osteoporosis in men with prostate carcinoma. Cancer, 97: 789,
2003
3. Dhillon, T. and Waxman, J.: Osteoporosis and prostate cancer.
Br J Cancer, 89: 779, 2003
Relative Excess
Risk*
Person-Yrs
at Risk
Rate/10,000
Person-Yrs
10
31
64
141
177
181
2,746
5,700
11,229
15,602
13,755
8,344
36
54
57
90
129
217
2.4
5.3
5.4
3.4
3.3
2.4
13
22
57
106
128
165
2,745
5,714
11,252
15,672
13,856
8,399
47
38
51
68
92
196
5.3
4.4
4.6
5.1
2.8
4.1
4. Gooren, I. J., Lips, P. and Gijs, L.: Osteoporosis and androgendepleting drugs in sex offenders. Lancet, 357: 1208, 2001
5. Morote, J., Martinez, E., Trilla, E., Esquena, S., Abascal, J. M.,
Encabo, G. et al: Osteoporosis during continuous androgen
deprivation: influence of the modality and length of treatment.
Eur Urol, 44: 661, 2003
6. Melton, L. J., 3rd, Alothman, K. I., Khosla, S., Achenbach, S. J.,
Oberg, A. L. and Zincke, H.: Fracture risk following bilateral
orchiectomy. J Urol, 169: 1747, 2003
7. Dahmani, L., Wagner, B., Auzanneau, C., Irani, J. and Dore, B.:
Prevalence of osteoporotic fractures in patients treated by
androgen blockade for prostate cancer. Prog Urol, 13: 73, 2003
8. Hatano, T., Oishi, Y., Furuta, A., Iwamuro, S. and Tashiro, K.:
Incidence of bone fracture in patients receiving luteinizing
hormone-releasing hormone agonists for prostate cancer. BJU
Int, 86: 449, 2000
9. Daniell, H. W.: Osteoporosis after orchiectomy for prostate cancer. J Urol, 157: 439, 1997
10. Townsend, M. F., Sanders, W. H., Northway, R. O. and Graham,
S. D., Jr.: Bone fractures associated with luteinizing hormonereleasing agonists used in the treatment of prostate carcinoma. Cancer, 79: 545, 1997
11. Oefelein, M. G., Ricchuiti, V., Conrad, W., Seftel, A., Bodner, D.,
Goldman, H. et al: Skeletal fracture associated with androgen
suppression induced osteoporosis: the clinical incidence and
risk factors for patients with prostate cancer. J Urol, 166:
1724, 2001
12. Population Registration in Sweden. Brochure RSV 711B. National Tax Board, 2000. Available at http://skatteverket.se/
broschyrer/711b/711b03.pdf. Accessed June 11, 2004
13. Cancer Incidence in Sweden 2000. Article 2002-42-5, National
Board of Health and Welfare, Centre for Epidemiology, 2002.
Available at http://www.sos.se/fulltext/42/2002-42-5/2002-425.pdf. Accessed June 11, 2004
14. In-Patient Diseases in Sweden 1987–1996. Article 1999-42-004,
National Board of Health and Welfare, Centre for Epidemiology, 1999. Available at http://www.sos.se/epc/english/ParEng.htm. Accessed June 11, 2004
15. Suissa, S.: Relative excess risk: an alternative measure of comparitive risk. Am J Epidemiol, 150: 279, 1999
16. Johansson, J. E., Holmberg, L., Johansson, S., Bergstrom, R. and
Adami, H. O.: Fifteen-year survival in prostate cancer. A prospective, population-based study in Sweden. JAMA, 277: 467,
1997
17. Khosla, S., Melton, L. J., 3rd and Riggs, B. L.: Estrogens and
bone health in men. Calcif Tissue Int, 69: 189, 2001
18. Michaelsson, K., Lithell, H., Vessby, B. and Melhus, H.: Serum
retinol levels and the risk of fracture. N Engl J Med, 348: 287,
2003
19. Smith, M. R.: Bisphosphonates to prevent osteoporosis in men
receiving androgen deprivation therapy for prostate cancer.
Drugs Aging, 20: 175, 2003